Web standards and its implications on future compatibilityThursday, April 17, 2003
Via Zeldman, eVolt have published an article by Peter-Paul Koch titled "Forward compatibility and web standards". I am a little perturbed by the brash statement "All current standards-compatible browsers also support old-fashioned tag soup, and they will always continue to support it."
That current PC-based browser support tag-soup is correct, yet there is a substantial overhead in doing so - and this substantial overhead is a barricade that prevents the Web from being the ubiquitous medium it is intended to be. In effect, invalid markup is an obstacle that needs to be overcome. For as long as browsers have to support tag-soup and tag-salad, it prevents future long term growth and accessibility of the World Wide Web.
How much smaller can Internet Explorer, Opera and Mozilla be without the limitation of supporting tag-soup or tag-salad? Looking at the Compaq IPaq and its many competitors, this "limited platform" is a growing market. Japan's tremendous growth in this industry is only an inkling of the future of the Internet as a pervasive medium.
How are bloated browsers going to survive and be of use in a handheld industry - its certainly not by insisting on 200Mb+ storage devices just for running one application; its certainly not by forcing expensive components which raises the price of these devices to unacceptable levels. Its in this very situation where web standards help. Valid XHTML with presentation done by CSS is typically much smaller and easier to parse using generalised XML tools. This means browsers don't need their own bloated parsers, but can rely on standard components to parse requested pages. This is a massive cost savings both in terms of product pricing, and in terms of resources required to run a browser.
The principle is simple, if it is too expensive for people to listen to your message, then it is unlikely you will be heard. Web standards ensures that the visitor cost of listening to your message is on average cheaper than a random collection of tags. What good is a browser if it can't run on the modern hardware the visitor is using? Or does Peter-Paul Koch quietly admit that these devices should have no right to access the Web, thereby locking the Web down as an academic curiosity with no future and no real world ambitions?
Jim Dabell makes a very interesting comment on the weakness of building for today - based around the four reasons to validate.